Intelligence Through Simulated Evolution: Forty Years of Evolutionary Programming

摘要: Genesis motivation prediction experiments pattern recognition and classification control system design extension of early evolutionary programming concepts competitive goal-seeking some implications diversification two-person gaming against nonminimax players coevolution, pursuit evasion modeling time series simulated ecosystems the nature intelligence sequence induction with deterministic automata revising extending routing problems comparing crossover, inversion, mutation specialisations finding structure in data self-adaptation evolving neural networks S-expression multiple interacting programs games other applications.